Infections in hepatic, biliary, and pancreatic surgery

Infections cause significant morbidity and mortality in patients undergoing hepatopancreatobiliary (HPB) surgery. In the era of more extensive resections in elderly comorbid patients with the greater use of perioperative chemoradiotherapy and biliary instrumentation, surgical-site infection (SSI) rates after HPB procedures can be as high as 20% to 40% with an added substantial risk of intraabdominal infections. Infection is associated with increased hospital stay, operative times, transfusions, blood loss, intensive care unit use, and readmission rates. In addition to these short-term sequelae, long-term sequelae include worse oncologic outcomes. This chapter will outline the range of infectious complications that may accompany resections of the liver, biliary tree, and pancreas with important surgery-specific risk factors discussed. Potential strategies for mitigating infection risk in the perioperative period are also discussed.

Risk factors for surgical-site infection

There are inherent protective host mechanisms within the liver and pancreas ( Table 11.1 ; see Chapter 10 ). Some of these are altered in the perioperative phase or from the pathology mandating intervention, such as bacterial colonization of bile from preoperative stenting or instrumentation of the hepatic inflow. ^, There are two important sources of risk regarding the development of a postoperative SSI. Both patient -specific factors and surgery -related factors combine to yield varying degrees of risk. Surgery-related risk factors are discussed specifically as they pertain to hepatic, pancreatic, and biliary operations.

TABLE 11.1

Normal Host Defense Mechanisms in the Hepatobiliary System and Pancreas

	HEPATOBILIARY	PANCREAS
Physical	Biliary sphincter	Pancreatic sphincter
	Hepatic tight junctions	Pancreatic tight junctions
	Bile flow	Pancreatic juice flow
	Mucus	Mucus
	Cilia	Cilia
Chemical	Bile salts	Pancreatic fluid
Immunologic	Kupffer cells	—
	Immunoglobulin A	Immunoglobulin A
	Fibronectin Complement	Complement

Patient-related risk factors

Patient-related risk factors may not always be modifiable at the time of operation, but it is important to be aware of them while the patient receives care. They include age, nutritional status, diabetes, smoking, obesity, coexisting infections at a remote body site, colonization with microorganisms, altered immune response, and length of preoperative stay. Interventions to modify these risk factors can be employed throughout the perioperative period from initial consultation until long-term follow-up ( Tables 11.2 to 11.4 ).

TABLE 11.2

General Preoperative Interventions to Prevent Surgical-Site Infection

Modified from Horan TC, Gaynes RP, Martone WJ, Jarvis W, Emori TG. CDC definitions of nosocomial surgical site infections: a modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol. 1992;13:606–608.

INTERVENTION	EVIDENCE	REFERENCES
Reduce hemoglobin A1c levels to <7% before operation	Class II data	Mangram et al. (1999)
Smoking cessation 30 days before operation	Class II data	Mangram et al. (1999)
Administer specialized nutritional supplements or enteral nutrition to patients at severe nutritional risk for 7-14 days preoperatively; preoperative parenteral nutrition should not be routinely used, except selectively in patients with severe underlying malnutrition	Class I and class II data with significant heterogeneity	Anonymous (1991) ; Mangram et al. (1999) ; Weimann et al. (2006)
Adequately treat preoperative infections, such as urinary tract infections	Class II data	Mangram et al. (1999)

TABLE 11.3

General Perioperative Interventions to Prevent Surgical-Site Infection

Modified from Kirby JP, Mazuski JE. Prevention of SSI. Surg Clin North Am. 2009;89(2):365–389, viii.

INTERVENTION	EVIDENCE	REFERENCES
Remove hair only if it will interfere with the operation; hair removal by clipping immediately before the operation or with depilatories; no preoperative or perioperative shaving of surgical site	Class I data	Anderson et al. (2008) ; Bratzler (2006) ; Kjønniksen et al. (2002) ; Mangram et al. (1999) ; Springer (2007)
Use an antiseptic surgical scrub or alcohol-based hand antiseptic for preoperative cleansing of the operative team members’ hands and forearms	Class II data	Anderson et al. (2008) ; Mangram et al. (1999)
Prepare the skin around the operative site with an appropriate antiseptic agent, including preparations based on alcohol, chlorhexidine, or iodine/iodophors	Class II data	Anderson et al. (2008) ; Digison (2007) ; Mangram et al. (1999)
Administer prophylactic antibiotics for most clean-contaminated and contaminated procedures, and selected clean procedures; use antibiotics appropriate for the potential pathogens	Strong class I data	Anonymous (1999) ; Bratzler (2006) ; Classen et al. (1992) ; Mangram et al. (1999) ; Springer (2007)
Administer prophylactic antibiotics within 1 hr before incision (2 hr for vancomycin and fluoroquinolones)	Strong class II data	Anonymous (1999) ; Bratzler (2006) ; Classen et al. (1992) ; Mangram et al. (1999) ; Springer (2007)
Use higher dosages of prophylactic antibiotics for morbidly obese patients	Limited class II data	Forse et al. (1989) ; Mangram et al. (1999)
Use vancomycin as a prophylactic agent only when there is a significant risk of MRSA infection	Class I data	Anderson et al. (2008) ; Anonymous (1999) ; Bolon et al. (2004) ; Finkelstein et al. (2002) ; Mangram et al. (1999)
Provide adequate ventilation, minimize operating room traffic, and clean instruments and surfaces with approved disinfectants	Class II and class III data	Anderson et al. (2008) ; Mangram et al. (1999)
Avoid “flash” sterilization	Class II data	Anderson et al. (2008) ; Mangram et al. (1999)
Carefully handle tissue, eradicate dead space, and adhere to standard principles of asepsis	Class III data	Anderson et al. (2008) ; Mangram et al. (1999)
Leave contaminated or dirty infected wounds open, with the possible exception of wounds following operations for perforated appendicitis	Limited class I, class II data	Brasel et al. (1997) ; Cohn et al. (2001) ; Mangram et al. (1999)
Redose prophylactic antibiotics with short half-lives intraoperatively if operation is prolonged (for cefazolin if operation >3 hr) or if there is extensive blood loss	Limited class I, class II data	Mangram et al. (1999) ; Scher (1997) ; Swoboda et al. (1996)
Maintain intraoperative normothermia	Class I data, some contradictory class II data	Anderson et al. (2008) ; Barone et al. (1999) ; Bratzler (2006) ; Mangram et al. (1999) ; Sessler & Akca (2002) ; Springer (2007) ; Walz et al. (2006)

MRSA, Methicillin-resistant Staphylococcus aureus .

TABLE 11.4

General Postoperative Interventions to Prevent Surgical-Site Infection

Modified from Kirby JP, Mazuski JE. Prevention of SSI. Surg Clin North Am. 2009;89(2):365–389, viii.

INTERVENTION	EVIDENCE	REFERENCES
Discontinue prophylactic antibiotics within 24 hours after the procedure (48 hr for cardiac surgery and liver transplant procedures); preferably, discontinue prophylactic antibiotics after skin closure	Class I data	Anonymous (1999) ; Bratzler (2006) ; DiPiro et al. (1986) ; Mangram et al. (1999) ; Springer (2007)
Maintain serum glucose levels <200 mg/dL on postoperative days 1 and 2	Class II data	Anderson et al. (2008) ; Bratzler (2006) ; Carr et al. (2005) ; Furnary et al. (1999) ; Lazar et al. (2004) ; Springer (2007) ; Zerr et al. (1997)
Monitor wound for the development of surgical-site infection	Class III data	Anderson et al. (2008) ; Mangram et al. (1999)

Surgery-specific risk factors

Hepatic resection (see Chapters 101 and 102 )

Preoperative risk mitigation

Hepatic resection removes Kupffer cell mass, which is the liver’s principal mechanism for clearing the portal inflow of enteric microorganisms and their associated toxins. Hepatic resection also decreases bile production with a consequent impairment of the chemical and immunologic effects of bile salts. Resection with biliary reconstruction or biliary stenting also bypasses the sphincter of Oddi and predisposes patients to bilioenteric reflux and cholangitis. Depending on the extent of resection, a normal healthy liver in a reasonable surgical candidate may be able to compensate for these changes. However, this may not be the case for the patient with diseased liver parenchyma. The preoperative mitigation of the risk of postoperative infectious complications after hepatic resection therefore begins with a full appreciation of the preexisting condition of the patient’s liver. Adjunctive assessments such as liver biopsy and measurement of portal venous pressures may be necessary when there is uncertainty concerning the health of the liver.

Yang and colleagues (2014) found cirrhosis and hepatolithiasis to be independent preoperative risk factors for the development of postoperative SSIs. Garwood and colleagues (2004) showed that the extent of hepatic resection, age, and comorbidity is associated with postoperative infectious complications. Schindl and colleagues (2005) established a relationship among the extent of resection, residual liver volume, and the development of infection. Although a precise residual liver volume to predict postoperative infection could not be found, there was a significant relationship linking severe hepatic dysfunction and postoperative infection (see Chapter 102 ). Furthermore, severe hepatic dysfunction could be predicted by small residual liver volume and high body mass index (BMI). Nanashima and colleagues (2014) similarly demonstrated that liver failure was significantly associated with deep SSIs. A recent study has similarly proposed that the future liver remnant (FLR) should be more than 45% in patients older than 69 to minimize postoperative complications including sepsis.

If the risk of postoperative hepatic dysfunction is deemed too high for formal resection, then other treatment modalities may be needed. For example, parenchymal-sparing techniques, such as segmental hepatectomy (see Chapter 102B ), ablation (see Chapters 95 and 96 ), or arterial-based modalities (see Chapters 94 , 97 , and 100 ) may be required. Preoperative portal vein embolization can be considered in certain patients in whom the FLR volume is too low and/or of poor quality. There is also increasing enthusiasm in facilitating resection by using hepatic vein embolization or ALPPS (see Chapters 102C and 102D ). When the risk of postoperative complications is prohibitive, then not operating or ablating may be the prudent course of action. It should, however, be noted that nonoperative strategies such as yttrium-90 radioembolization are also associated with infectious complications, and efficacy for many of these modalities has yet to be established.

The use of systemic chemotherapy is increasingly common in the overall treatment plan for patients undergoing hepatic resection, especially in patients with colorectal liver metastases (see Chapters 50 and 97, 98, 99 ). Neoadjuvant chemotherapy may increase the risk of infection due to its negative effects on the liver, including steatosis, steatohepatitis, and sinusoidal obstruction syndrome (see Chapter 69 ). It is, however, an important part of multimodal management of hepatic tumors, especially colorectal liver metastases, and confers superior oncologic outcome. Scilletta and colleagues (2014) suggested that neoadjuvant chemotherapy was not a significant risk factor for SSIs in patients undergoing liver resection for colorectal hepatic metastases.

Nordlinger and colleagues (2008) conducted a randomized controlled trial comparing liver resection for resectable colorectal liver metastases in patients with and without perioperative chemotherapy. Perioperative chemotherapy was defined as six cycles of 5-fluororuracil plus leucovorin and oxaliplatin (FOLFOX4) before and after surgery. There were 182 patients within each arm of the study. Infectious complications that were analyzed included wound, intraabdominal, and urinary infections. There was a trend toward higher rates of these complications in the perioperative chemotherapy group, but it was not statistically significant. In the setting of resectable colorectal cancer liver metastasis, chemotherapy before hepatectomy is safe from a postoperative infectious standpoint and remains indicated overall to achieve optimal oncologic outcomes. Liver-directed chemotherapy via hepatic artery infusion pumps also does not seem to increase the risk of infectious complications (see Chapters 69 , 97 , and 98 ).

Preoperative nutrition also requires careful consideration. A recent review by Walcott-Sapp et al. (2018) has expanded on the importance of recognizing malnutrition before hospital admission and considering oral supplementation with high-calorie, high-protein drinks and avoiding unnecessary preoperative fasting in the immediate preoperative phase. A recent trial by Russell et al. (2019) also evaluated the value of immunonutrition in patients undergoing elective hepatectomy, but the majority of patients did not have malnutrition and there were no differences in outcome between the two groups (see Chapters 26 and 27 ).

Other preoperative contributors to postoperative infectious complications after hepatic resection include advanced age, presence of diabetes mellitus, obesity, presence of an open wound, hypernatremia, hypoalbuminemia, elevated serum bilirubin, dialysis, comorbid conditions, repeat hepatectomy, and hepatic steatosis. ^, Because many of these preexisting conditions may not be modifiable before the time of operation, any resection must be considered in light of the general condition of the patient.

Operative risk mitigation

There are several operative risk factors that are associated with postoperative infectious complications. These include bile leaks, surgery duration, increased blood loss, and iatrogenic bowel perforation. ^, ^, ^, It is thus important to mitigate these factors; one deliberate strategy can be intraoperative identification and treatment of bile leaks. Numerous strategies have been proposed for this including the use of indocyanine green, methylene blue, hydrogen peroxide, or air (see Chapters 24 and 25 ).

An intraoperative air leak test has been shown to be effective in the detection of bile leaks, thus, decreasing the rate of postoperative biliary complications. This maneuver involves the placement of a transcystic catheter that is used to inject air into the biliary tree after the upper abdomen is submerged in saline and the distal common bile duct is occluded. Bile leaks are identified at the site of streaming air bubbles and are directly repaired. The authors compared the rates of postoperative biliary complications among 103 patients who underwent air-leak testing and 120 matched patients who underwent hepatic resection before air-leak testing was used. None of the hepatic resections in either group were accompanied by biliary reconstruction. The authors noted a significantly lower rate of postoperative bile leaks in the air leak–tested group (1.9% vs. 10.8%, P = .001).

Minimally invasive approaches (laparoscopic/robotic) have also reported lower rates of wound infections and lower rates of pulmonary complications with comparable rates of bile leak and noninferior oncologic outcomes. ^, Although laparoscopic liver resections are commonly used for minor hepatectomy, major hepatectomy is also increasingly performed using minimally invasive techniques.

Regarding parenchymal transection techniques, no one method or combination of methods has been shown to be convincingly superior, although a recent network meta-analysis has suggested (with significant caveats) that energy devices may be best in reducing overall complications (see Chapter 118 ). It is thus recommended that the surgeon use the technique that is most familiar, while limiting the amount of necrotic liver parenchyma left behind. It is also important to suction any pooled blood and bile at the end of the operation.

Perioperative antibiotics

A reasonable approach to the prevention of SSI in the setting of hepatic resection would be to administer antimicrobial prophylaxis for all elective hepatic resections regardless of anticipated bilioenteric reconstruction. The recommended antimicrobial agent for biliary tract procedures from the Clinical Practice Guidelines for Antimicrobial Prophylaxis ( https://www.ashp.org ) is cefazolin. This online resource is frequently updated and can be used to guide antibiotic choice in line with institutional considerations. Other options include cefoxitin because it is a single agent with broad coverage. Whatever agent is used, it should be given within 60 minutes of skin incision and re-administered appropriately intraoperatively to maintain adequate tissue levels. There is no evidence to support routine use of postoperative antibiotics.

Drains

The rationale for leaving an intraabdominal drain is to detect and prevent biloma formation in the event of bile leakage after hepatic resection (see Chapters 28 and 118 ). Bile leakage and subsequent biloma formation is an important contributor to infectious complications after hepatic resection. However, the overall trend reflected in the literature does not support the routine use of drains in elective hepatic surgery. Foregoing prophylactic drainage after elective hepatic surgery is consistent with the general notion that drainage may be unnecessary in most gastrointestinal (GI) operations. Petrowsky and colleagues (2004) studied the value of prophylactic drainage in GI surgery in a systematic review and meta-analysis, concluding that many GI operations can be safely performed without the use of drains. Regarding liver surgery specifically, this article suggests that surgical drains do not necessarily prevent biloma formation and do not always prevent the need for percutaneous drainage. A grade A recommendation was given against prophylactic drainage in elective hepatic resection. This is supported by several randomized studies and a systematic review. A more recent multicenter international prospective study also concluded that intraoperatively placed surgical drains do not prevent the need for additional percutaneous drainage. Another study evaluating hepatectomy within the United States showed that routine drain placement did not prevent nor diagnose bile leaks and was associated with more interventions, a higher rate of readmissions, and longer length of stay. However, practical considerations such as urgent access to image-guided drain insertion postoperatively if required often affect clinical decision making and individual high-risk cases (e.g., central resections, complex nonanatomical resections, and high bilioenteric anastomosis) may warrant prophylactic drain placement. A reasonable strategy may be to avoid routine drain placement unless specific concerns exist and remove drains as early as possible postoperatively.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here